5022
J . Am. Chem. SOC 1987, 109, 5022-5023
kcal/M),6 and C O on Ni (3.3-4.2 kcal/M)I2 by using other desorption techniques. The selectivity and reactivity of molecular rearrangements on Pt surfaces have been found to be strongly dependent on the surface structure.s We have also observed this phenomenon: the Pt foil used in the above experiments was annealed to 800 OC before each measurement. Surfaces that were roughened by ion sputtering exhibited a different behavior. Details of the decomposition of both C6F6and C2F4 on Pt surfaces are presently under investigation by DRS and Auger, X-ray, and UV electron spectroscopy.
Q
E
l
Acknowledgment. This material is based on work supported by the National Science Foundation under Grant no. CHE8513966 and by the R. A. Welch Foundation under Grant no. E656. 0
(12) Mullins, D. A.; Roop, B.; White, J. M. Chem. Phys. Lett. 1986, 129, 511-515.
Electroreduction of CO to CH4 and CzH4at a Copper Electrode in Aqueous Solutions at Ambient Temperature and Pressure Yoshio Hori,* Akira Murata, Ryutaro Takahashi, and Shin Suzuki Department of Synthetic Chemistry Faculty of Engineering, Chiba University Yayoi-cho. Chiba 260, Japan Received February 17, 1987 Carbon monoxide is a substance of primary importance from the viewpoint of utilization of carbon resources. Hydrogenation of CO in the gas phase has been widely studied, but only a small number of papers has been published about electrochemical reduction. Electroreduction of C O did not proceed effectively according to the previous papers;' the cathodic partial currents for electroreduction of C O were very small, not exceeding 5 X A cm-2. We briefly describe here an electroreduction of C O at a Cu cathode in aqueous solutions. This procedure allows an effective cathodic reduction of C O to form hydrocarbons and alcohols with appreciable current densities. An electrodeposited copper sheet (purity 99.999%), donated by Sumitomo Metal Mining Co. Ltd., was cut into an electrode (20 X 20 X 1 mm) with a copper lead strip attached. We did not mount the electrode in resin in order to avoid possible contamination of the boundary interface of electrode/resin. The electrode was polished with fine emery paper (no. 1500), electrolytically polished in 85% phosphoric acid for ca. 2 min at ambient temperature, and then rinsed with doubly distilled deionized water. Electrochemical measurements were conducted a t 18 "C with a three-compartment cell in which two anode compartments faced each side of the Cu electrode. The cathode compartment (36-mm inner diameter) was separated from the two anodes with sheets of cation exchange membrane (Selemion). The potential of the cathode was measured with respect to an Ag/AgCl reference electrode. The electrode potential was corrected for the IR drop between the Luggin capillary tip and the cathode. The catholyte (60 mL) was prepared from doubly distilled deionized water and reagent grade chemicals. The ca(1) (a) Petrova, G. N.; Efimov, 0. N.; Strelets, V. V. Izu. Akad. Nauk SSSR. Ser. Khim. 1982, 2608-2610. (b) Uribe, F. A,; Sharp, P. R.; Bard, A. J. J . Electroanal. Chem. 1983, 152, 173-182. (c) Yoneyama, H.; Wakamoto, K.; Hatanaka, N.; Tamura, H . Chem. Lett. 1985, 539-542. (d) Yamamura, S.; Kojima, H.; Kawai, W. J . Electroanal. Chem. 1985, 186, 309-312. (e) Ogura, K.; Takagi, M. Ibid. 1985, 195, 357-362. (f)Kolyagin, G. A,; Danilov, V. G.; Kornienko, V. L. Elektrokhimiya 1985,21, 133-135.
0002-7863/87/l509-5022$01.50/0
I -0.4
-0.6 -0.8
-10
-1.2
vs S H E ) Figure 1. Voltammograms obtained in a phosphate buffer solution (0.17 M KH,P04/0.03 M K,HP04, p H 6.1) saturated with N2or CO. Scan rate 50 mV s-'. Potential ( V
tholyte was purified by preelectrolysis with a 30 X 20 mm pt black A cm-2 under purified N, atmosphere for cathode at 2.5 X more than 15 h. The Pt black cathode was regenerated by anodic polarization in 0.5 mol dm-, H,S04 before use. Carbon monoxide (high purity grade, >99.95%, hydrocarbon content C0.5 ppm, dew point below -65 "C), purchased from Nihon Sanso Co. Ltd., was introduced to the catholyte after passing through a KOH solution. Voltammetric measurements were conducted with electrolytes saturated with purified C O or N,. Coulometric measurements at constant currents were conducted in KHCO,, KOH, and a phosphate buffer aqueous solution with C O bubbled into the catholyte (flow rate: ca. 70 mL min-') for 30 min; the catholyte was stirred vigorously with a magnetic stirrer. CH4 and C2H4in the effluent gas from the cell were analyzed at 5-min intervals by a gas chromatograph equipped with an FID detector (Shimadzu 3BF). H2 was analyzed by a gas chromatograph equipped with a TCD detector (Shimadzu 3AH). The volume of the gas sampling tube was 3 mL. The analytical data were averaged over 30 min of electrolysis. The limit of quantitative analysis was 0.6 ppm for CH4 and 0.3 ppm for C2H4;these values correspond to a faradaic yield of 0.1% in a constant current (2.5 mA cm-2) electrolysis. C2HjOH and n-C,H,OH in the solution were also analyzed after electrolysis by a gas chromatograph (Shimadzu 3BF). Formaldehyde in the solutions was determined by the chromotropic acid colorimetric method. The limit of quantitative analysis for the soluble products was 3 X mol dm-3 for C2HjOH and n-C,H,OH and 0.1 X 10-6 mol dm-, for HCHO. The faradaic yields for these limits are 0.3% for C2H50H and n-C3H,0H and 0.002% for HCHO. The following analytical columns were employed for the identification of the products by the gas chromatograph; molecular sieve 5A and 13X, Porapak N Porapak Q, activated alumina TR, and silica gel for CH4; Porapak N + Porapak Q, activated alumina TR, and silica gel for CzH4; Porapak Q, PEG 400/Chromosorb, Unisole 30T (alkylene glycol phthalate)/Chromosorb, and PEG 20M/Shimalite for C2HjOH; PEG 400/Chromosorb, Unisole 30T/Chromosorb, and PEG ZOM/Shimalite for n-C3H70H. The formation of CHI and C2H4 was also confirmed by a gas chromatograph mass spectrometer. Figure 1 shows voltammograms obtained for N2 and C O saturated phosphate buffer solutions (pH 6.1). The cathodic current is greatly suppressed in the presence of CO. Anodic oxidation of CH,OH, as is well known, is markedly interfered with by some intermediate species formed in the reaction, e.g., by C O adsorbed on the electrode or some other related species.2 Adsorption of
+
(2) (a) Shimazu, K.; Kita, H. Denki Kuguku 1985, 53, 652-662. (b) Beden, B.; Lamy, C.; Bewick, A,; Kunimatsu, K. J . Electroanal. Chem. 1981, 121, 343-347. Kunimatsu, K. J . Electroanal. Chem. 1982, 140, 205-210.
0 1987 American Chemical Society
5023
J . Am. Chem. SOC.1987, 109, 5023-5025 Table I. Faradaic Efficiencies of Various Products from the Electroreduction of C O at a C u Electrode in Aqueous Solutions electrolyte
PH
potential vs. S H E
CHI
CZH4
EtOH
faradaic efficiency/% n-PrOH HCHO
HZ
total
Current 1.0 mA cm-2 phosphate"
KHCO,~
6.1 9.4
-1.10 -1.29
0.3
6.1 9.6
-1.18 -1.40
9.3 16.3
6.0 9.6 12.9
-1.21 -1.45 -1.47
16.8 16.2
1.1
1.7 21.6
0.0 1.3
0.0 2.0
0.2 0.04
91.4 63.3
93.6 89.3
0.0 1.5
0.2 0.1
81.1 45.5
92.9 95.5
0.0 0.3 1.1
0.02 0.03 0.05
75.4 65.4 70.7
93.9 90.1 92.8
Current 2.5 mA cm-2 phosphatea
KHCO:
2.3 21.2
0.0 10.9
Current 5.0 mA cm-2 phosphate"
KHCO: KOHC
1.O
1.7 5.5 14.1
0.0 2.7 5.8
"Phosphate buffer solution containing KH,PO4 (0.17mol dm-') and K2HP0, (0.03mol dm-'). 'KHCO, (0.1 mol dm-'). cKOH (0.1mol dm-').
C O or of other molecules may prevent the H, evolution in the present system as well. A small hump appears at -0.86 V vs. SHE in the voltammogram obtained for CO-saturated solution. A coulometric measurement at a controlled electrode potential of -0.86 V vs. S H E yielded only H2, and the faradaic efficiency for Hz formation was more than 97%. The source of the hump is not clarified yet at the present stage. The gaseous products, CH4, C2H4,and H2, appeared in the effluent gas after the electrolysis started; the concentrations remained virtually constant during the electrolysis. C2H6was not detected in the effluent gas. CH4, C2H4, and H2were not detected in the effluent gas (CO) within the limit of detection before the electrolysis started. The products contained in the electrolytes were C 2 H 5 0 H ,n-C3H70H,and HCHO. C H 3 0 H was not detected in the solution. These soluble products were not found within the limit of detection in a blank experiment carried out under exactly the same conditions without electrolysis. Table I presents the faradaic efficiencies of products obtained in various coulometric measurements. The faradaic efficiencies were calculated on the basis of the number of electrons required for the formation of one molecule of the products from C O and H2O: six for CH4, eight for CzH4, eight for C2H50H,12 for n-C3H70H, two for HCHO, and two for Ha. The total values of the faradaic efficiency exceed 90%. Hence the tabulated substances are the major products of the reaction. The partial current for CO reduction (I,) exceeds 1 mA cm-' for KHCO, solution at -1.40 V vs. S H E with the total electrolytic current of 2.5 mA ern-,. The overvoltage of the reaction is large (EO(COICH4) = -0.31 V, Eo(CO/C2H4) = -0.40 V VS. S H E at pH 9.6).' However, the value Z, obtained in this study is much higher than any values reported in the previous papers, Le., less than 5 X mA This I , appears to be of the same order of magnitude of the limiting current due to the transport of C O to the electrode, since the solubility of C O is low in water. Table I shows some remarkable features of this reaction. The production of C2H4 prevails over that of CH4 at less negative potentials (less total current) and at higher pH solutions. The increased formations of CzH50Hand n-C3H70H are accompanied with high faradaic efficiencies of C2H4. Hence, CzH4, C2H50H, and n-C,H,OH will be formed via some common intermediate species. H 2 formation is markedly interfered with C O or CO related species adsorbed on the electrode, as indicated in Figure 1. These adsorbed species will probably be reduced at more negative potentials to yield hydrocarbons and alcohols. The Cu electrode, etched in 6% HNO, aqueous solution instead of by electropolishing, was also tested for the cathode. Electrolysis with this electrode was conducted with 0.1 M KHCO, at 2.5 mA cm-*. The current efficiencies for the products were as follows: CH4, 20.4%; C2H4,8.3%; C2H50H, 3.7%; n-C,H,OH, 0.14%; HCHO, 0.02%; H,, 63.4%. These values are close but not exactly equal to those obtained with an electropolished electrode (Table I). The reason is unknown at present. ~
We previously reported the production of CH4 and C2H4 in electrochemical reduction of C 0 2 at a Cu electrode in aqueous KHCO, s o l ~ t i o n . ~The present results suggest that the electroreduction of COz may proceed with C O or CO-derived intermediates formed at the Cu electrode. __
~
(4)Hori, Y.;Kikuchi, K.; Murata, A,; Suzuki, S. Chem. Lett. 1986, 897-898.
Platinacyclobutanes on the Route to Cyclopropanation T. W. Hanks and P. W. Jennings* Gaines Hall, Department of Chemistry Montana State University Bozeman, Montana 5971 7 Received March 2, 1987 Recently we reported on the preparation and bonding character of platinum complexes arising from the reaction of platinum(I1) with diazoderivatives (eq l).' It was clear from these results that CI
CI
I
II
1 equiv. Py
II-)
R= H,H H,CH, and H, C0,Et ~
Pt(I1) cannot provide adequate orbital overlap with the electrophilic carbon * to form a K bond, and pyridine is required for stabilization. In this communication, results are reported on a similar reaction with use of diazofluorene which, in contrast to those cited above, can provide substantial stabilization of the electrophilic center by electron donation from the aryl moieties. These results not only impact on the mechanism of metal-facilitated cyclopropanations but also represent the first example of a platinum-catalyzed cyclopropanation. The results are shown in Scheme I. To prepare le, Zeise's dimer was reacted with 1 equiv of pyridine in chloroform yielding l a in quantitative yield. Subsequent reaction with diazofluorene in chloroform containing a second equivalent of pyridine gave the bright red complex l e in 94% yield. It was further purified by trituration with pentane and silica gel chromatography. From its I'C, 'H, and 195PtNMR spectra, structure l e was readily established. The data are as = 376 Hz, 2 C), 60.9 follows: I3C (CDCI,, ppm) -4.65 (t, lJpt,C (s, 2Jpt,c = 95 Hz, 1 C) and nine aromatic resonances at 153.4 (d, 2 C), 152.3 (s, 2 C), 149.3 (d, 4 C), 141.0 (s, 2 C), 138.1 (d, 2 C), 129.0 (d, 2 C), 125 (d, 4 C), 119.7 (d, 2 C), and 118.7 (d, 2 C) accounting for all 22 aromatic carbons. The resonances at
~~~~~
(3) Estimated from thermodynamic data. Lange's Handbook of Chemistry, 12th ed.; Dean, J . A,, Ed.; McGraw-Hill, New York, 1979.
0002-7863/87/1509-5023$01 SO10
(1) Hanks, T. W.; Jennings, P. W. Organometallics
0 1987 American Chemical Society
1987,9, 28.
